Hey all! I’m new here but quickly wanted to ask a question regarding a particular issue I am facing. I have imaging data from PCASL with 5 PLD blocks and mcflirt picks up the static tissue signal (due to reduced background suppression) from the volumes of my last pld block as motion and thus there seems to be a motion artifact for the jump.
Also, I have been running mcflirt in my oxford_asl pipeline and I’m getting abnormally low values for perfusion but when i removed the mcflirt command and ran it, I’ve been getting somewhat normal values. Can someone help out or suggest something.
This was an issue in ASLPrep recently as well. Switching the MCFLIRT cost function from normcorr to mutualinfo seems to have solved the problem for us.
that did not work unfortunately. would you recommend maybe normalizing the last volumes to kind of suppress the static tissue signal?
The only other thing ASLPrep does is it separately applies MCFLIRT to each volume type, though the reference image is the same.
One thing we’ve discussed implementing, but haven’t needed to, is looping over each of the volumes in the run and using ANTs to register them to the reference image, rather than using MCFLIRT or 3dVolReg.